Evidence of pervasive trans-tensional deformation in the northwestern Wharton Basin in the 2012 earthquakes rupture area

HAL Id: insu-02282789

https://hal-insu.archives-ouvertes.fr/insu-02282789

Submitted on 10 Sep 2019

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Evidence of pervasive trans-tensional deformation in the

northwestern Wharton Basin in the 2012 earthquakes

rupture area

Nugroho Hananto, Asmoune Boudarine, Hélène Carton, Satish Singh,

Praditya Avianto, Jerome Dyment, Yanfang Qin, Dibakar Ghosal, Rina

Zuraida, Paul Tapponnier, et al.

To cite this version:

Nugroho Hananto, Asmoune Boudarine, Hélène Carton, Satish Singh, Praditya Avianto, et al..

Ev-idence of pervasive trans-tensional deformation in the northwestern Wharton Basin in the 2012

earthquakes rupture area. Earth and Planetary Science Letters, Elsevier, 2018, 502, pp.174-186.

�10.1016/j.epsl.2018.09.007�. �insu-02282789�

Contents lists available atScienceDirect

Earth

and

Planetary

Science

Letters

www.elsevier.com/locate/epsl

Evidence

of

pervasive

trans-tensional

deformation

in

the

northwestern

Wharton

Basin

in

the

2012

earthquakes

rupture

area

Nugroho Hananto

a

,

Asmoune Boudarine

b

,

Hélène Carton

b

,

Satish

C. Singh

b

,

e

,

∗

,

Praditya Avianto

a

,

Jérôme Dyment

b

,

Yanfang Qin

b

,

Dibakar Ghosal

c

,

Rina Zuraida

d

,

Paul E. Tapponnier

e

,

Christine Deplus

b

,

Kerry Sieh

e

a_{ResearchCenterforOceanography,IndonesianInstituteofSciences,JlPasirPutih1AncolTimur,JakartaUtara14430,Indonesia}

b_{EquipedeGéosciencesMarines,InstitutdePhysiqueduGlobedeParis(CNRS,ParisDiderot,SorbonneParisCité),1 rueJussieu,75238ParisCedex05,France} c_{IndianInstituteofTechnology,Kanpur,India}

d_{MarineGeologicalInstitute,JlDR.Junjunan236,Bandung,Indonesia}

e_{EarthObservatoryofSingapore,NanyangTechnologicalUniversity,N2-01A-XX,50NanyangAvenue,Singapore639798,Singapore}

a

r

t

i

c

l

e

i

n

f

o

a

b

s

t

r

a

c

t

Articlehistory:

Received26November2017

Receivedinrevisedform31August2018 Accepted5September2018

Availableonline18September2018 Editor:J.-P.Avouac

Keywords:

the2012IndianOceanearthquake diffusedeformation

WhartonBasin platebending shearzones normalfaults

The Wharton Basin in the Indian Ocean is one ofthe most extensively deformingocean basins,as confirmedbytheoccurrenceofseveralverylargeearthquakesstartingfromJanuary12,2012withMw 7.2followedbythegreatearthquakesofApril11,2012withMw8.6andMw8.2.AlthoughtheMw7.2 andMw8.2earthquakesseemtohaverupturedthere-activatedN–Sstrikingfracturezones,thelargest event(Mw8.6)requiredtherupturingofseveralfaults,obliquetoeachother,inaverycomplexmanner. In order to understand the nature of deformation in these earthquakes rupture zones, we recently acquired 90 000 km2 _{of bathymetry, 11 400 km of sub-bottom profiling, gravity and magnetic data}

covering theruptureareasofthe2012earthquakeseastoftheNinety-EastRidge,inthenorthwestern WhartonBasin.ThesenewdatarevealsixN8◦Estrikingre-activatedfracturezones(F5b,F6a,f6b,F7a, F7bandF8),wherethefracturezoneF6acanbefollowedforover400 kmandseemstobemostactive. TheepicentersoftheMw8.6andMw8.2earthquakeslieonthefracturezonesF6aandF7b,respectively. ThenewlyobservedfractureF5bintheeastisshort,andhasanextensionalbasinatitssoutherntip.The fracturezoneF8definestheeasternboundaryoftheNinety-EastRidge.Thepresenceofenechelonfaults and pull-apartbasinsindicateleft-lateral motionalongthesefracturezones.Inbetweenthesefracture zones, weobservepervasive290◦strikingright-lateralshearzonesat4–8 kmintervals;one ofwhich hascutthroughaseamountthatmighthaverupturedduringtheMw8.6earthquake.Wealsoobserve anotherN20◦Estrikingleft-lateralshearzonesinthevicinityofF7bandF8,whichiscoincidentwiththe strikeofoneofthenodalplanesoftheMw8.6focalmechanism.TheseN20◦Estrikingshearzonesare interpretedasRRiedelshearsandtheN290◦EstrikingshearzonesasRRiedelshears.Theseshearzones areformedbyaseriesofN335◦Estrikingenechelonnormalfaults.Ourdataalsoshowthepresenceof N65◦EstrikingthrustfaultseastoftheNinety-EastRidge,orthogonaltotheregionalprincipaldirection ofcompression. Furthermore, extensivebending-relatedfaultingisalsoobserved closetothe Sumatra trenchwithnormalfaultsalsostrikingatN335◦E,similartothenormalfaultsthatformtheshearzones. NormalfaultswithasimilarorientationarealsopresentatthesoutherntipofF5b.Weexplainallthese observationswithasinglecoherentmodelofdeformationintheWhartonBasin,whereadominantpart oftheregionalNW–SEcompressionalstressisaccommodatedalongtheN8◦Ere-activatedfracturezones, andtherestisdistributedalongshearzones,thrustandnormalfaultsbetweenthesefracturezones.The thrustandnormalfaultsareorthogonaltoeachotheranddefinethedirectionofprincipalcompressive andextensivestressesintheregionwhereasthetwoshearzonesystemsformaconjugatepair.

©2018TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

*

Correspondingauthorat:EquipedeGéosciencesMarines,InstitutdePhysique duGlobedeParis(CNRS,ParisDiderot,SorbonneParisCité),1 rueJussieu,75238 ParisCedex05,France.

E-mailaddress:singh@ipgp.fr(S.C. Singh).

1. Introduction

The largest recorded intra-plateoceanicstrike-slip earthquake ofMw 8.6occurredonApril 11,2012intheWhartonBasinwith its epicenterabout120 kmsouthwestoftheSundatrench.Itwas

https://doi.org/10.1016/j.epsl.2018.09.007

N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 175

Fig. 1. (a)RegionaltectonicsettingofthecompositeIndo-Australianplate.DashedyellowlinesdefinethediffuseplateboundariesbetweenIN(India),AU(Australia)andCAP (Capricorn)plates(afterRoyerandGordon,1997).Redboxshowslocationoffigure(b).CL=Chagos–LaccadiveRidge;CIR=CentralIndianRidge;SEIR=SoutheastIndian Ridge;NER=Ninety-EastRidge.SeafloortopographyisfromETOPO1(AmanteandEakins,2009).Thebeachballsindicatenormal(green),thrust(black)andstrike-slip(red) earthquakes.Allthebeachballsareplottedwiththeircircleradiiproportionaltomagnitude.(b)OceanicplatestructureintheWhartonBasin.Fracturezones(reddashed lines)andfossilWhartonSpreadingCentre(doubleblackline)arefromSinghetal. (2011) identifiedfromaltimetrydata.Whitearrowsmarktheconvergencedirectionof theIndo-AustralianPlatewithrespecttotheSundaPlate(Prawirodirdjoetal.,2000).BeachballsshowhistoricalseismicityrecordedintheWhartonBasinsince2010.Focal mechanismsrepresentallmagnitudes>5GlobalCMTsolutions(Ekströmetal.,2012) fromJanuary2010toMay2017.OnlyMw>7.8 thrustearthquakesinthesubduction zoneareshown.Thecombinedred–greenarrowsindicatethedirectionsandtherelativemagnitudesoftheprincipaldeviatoriccompressionandtensioncomponentsof stress(GordonandHouseman,2015).ThethreeorangebeachballsindicatetheMw8.6mainshockanditsMw8.2aftershock,andMw7.2foreshock.The2ndMarch2016 Mw7.8earthquakeisalsomarked.ThewhitelinedelimitsthenewlyacquiredbathymetricdatashowninFig.3.(Forinterpretationofthecolorsinthefigure(s),thereader isreferredtothewebversionofthisarticle.)

preceded by a foreshock of Mw

=

7.2 on January 10, 2012 and

followedbyan aftershock ofMw

=

8.2twohours later(Duputel etal.,2012), alongwithhundreds ofsmaller aftershocks(Figs.1, 2). It has been suggestedthat this earthquake occurredas a re-sultofstresstransferontheincomingplateafterthe2004Mw9.2

Sumatra–Andamanearthquake andthe 2005Mw 8.7Nias

earth-quake(Delescluseetal.,2012).

This part of the Wharton Basin lies in a broad deformation zone within the composite Indo-Australianplate, extendingfrom theChagos–Laccadive ridge inthe west to theInvestigator Ridge inthe east (RoyerandGordon, 1997) (Fig. 1a). Although this re-gioniswellknownfordistributeddeformationwithin anoceanic plate(Gordon, 2000),manydetails aboutits seafloormorphology andtectonicactivityarelacking,owingtoits largeextentand re-mote location. GPS and marine geophysical studies indicate that compressionalfaultsandfoldsoccurwestoftheNinety-EastRidge (NER)inthe CentralIndian Basinsouth ofIndia (Bulland Scrut-ton, 1990; Delescluse and Chamot-Rooke, 2007). In contrast, de-formation is mainly accommodated by strike-slip faulting along re-activatedN–S striking fracture zones in theWharton Basin to theeastoftheNER(Deplusetal.,1998; Deplus,2001) (Fig.1b).

The occurrenceof such a large earthquake witha high stress

drop away from major plate boundaries came as a surprise

(McGuireandBeroza,2012).Althoughtheforeshockandthemain aftershockseemtohaverupturedalongre-activatedfracturezones, themainevent(Mw8.6)appears tohaverupturedseveralfaults, oblique to one another, in a very complex manner (Hill et al.,

2015; Wei et al., 2013; Ishii et al., 2013; Meng et al., 2012; Yue et al., 2012) (Fig. 2). This suggests that deformation in the northern Wharton Basin is distributed over a number of faults, but the exact geometry of this fault system and the identifica-tion of the main structuresare still open questions. Most mod-els based on seismological and geodetic studies agree that the

Mw 8.6 main shockinvolved ruptureon one NNE–SSW trending

fault along with atleast one WNW–ESE trending fault, withthe

seismic moment released dominantly during the NNE–SSW

rup-ture. The limited existing bathymetry data show recent activity along thesefracturezones(Deplusetal., 1998; Graindorgeetal., 2008) but their strike is no more than 8◦E (Carton etal., 2014; Singhetal.,2017).Ontheother hand,theGCMT solutionforthe maineventshowsonenodalplanestriking20◦E,andnoneofthe faultgeometriesassociatedwithmodelsofcoseismicslip distribu-tionalignwiththeexistingfracturezones(Fig.2).Basedonlimited bathymetry(75 kmby100 km)andhigh-resolutionseismic reflec-tion data in the vicinity of the Mw 8.2 aftershock, Singh et al. (2017) foundtheexistenceofN294◦Etrendingshearzonesin ad-dition tothe N–S re-activatedfracture zones, andsuggestedthat thesetwofaultsystemsformaconjugatepairoffaults, accommo-datingthelarge-scaledeformationobservedintheWhartonBasin. However,the2012Mw8.6earthquakeruptureextendsoveravery largeareathatisverypoorlysampledbymarinegeophysicaldata, leadingtomajoruncertaintiesregardingtheexactlocationand ge-ometryoffaultsactivatedduringthisearthquake.

Fig. 2. 2012WhartonBasinearthquakesequence(orangebeachballs).The epicen-tersforthefirst4daysofthe2012aftershocksequenceareshownwithyellowdots (USGScatalog).Thecolored linesindicatefivemodelsofcoseismicslipdistribution ofthe Mw8.6earthquake (Hilletal., 2015; Weiet al.,2013; Ishiiet al.,2013; Mengetal.,2012; Yueetal.,2012).ThereddashedlinesindicatefracturezonesF4 toF8.NER:Ninety-EastRidge.

In July 2016, we acquired

∼

90 000 km2 of high-resolution multibeambathymetry, 11 400 kmof3.5 kHzecho-sounder, grav-ityandmagneticdataonboardtheFrenchResearchVesselMarion Dufresne,duringtheMIRAGE(MarineInvestigationoftheRupture

Anatomyofthe 2012Great Earthquake)experiment coveringthe

2012 earthquakes rupture areas in the Wharton Basin. Here we

presentthefirstdetailedinvestigationoftheseafloormorphology and its relations with the Mw 8.6 earthquake rupture, and pro-videacomprehensivemodelofactivedeformationintheWharton Basin.

2. Tectonicsetting

The NE Indian Ocean is a complex region with a rich

tec-tonichistoryinitiatedduringthebreakupofIndiaandAustraliaat

∼

155 Ma(Sageretal.,1992; HeineandMüller,2005).The spread-ing between India andAustralia was initially taking place along a NW–SE direction, creating the western Australian basin. It re-orientedalong anE–W directionat

∼

100 Ma,forming themajor N–S orientedfracture zones observed inthe Wharton Basin. The spreadingvelocitywasfastpriortothecollisionofIndiawith Eura-sia,withIndiamoving at

∼

8–10 cm/yrat

∼

70 Ma, andabruptly

slowing down to 2–4 cm/yr at

∼

40 Ma (Molnar and

Tappon-nier,1977),whilethespreadingattheWhartonSpreadingCentre ceasedbetween38 Maand36.5 Ma(Jacobetal.,2014).Following thiscessation,IndiaandAustraliabecamea singlerigidplate(Liu etal.,1983).

Royer and Gordon (1997) suggested that the Indo-Australian platecanbeconsideredacompositeplatesystemwiththree com-ponent plates(India, AustraliaandCapricorn;Fig. 1) boundedby

diffuse boundary regions. Deformation within the Indian Ocean

extendsfromtheCentralIndianBasinacrosstheNERtothe north-west shelf ofAustralia (Fig. 1a)(Weissel etal., 1980; Petroy and Wiens, 1989; Wiens and Stein, 1984). The northern side of the

composite Indo-Australian plate displays a change in boundary

conditions from west to east, with continental collision of India withEurasia,obliquesubductionattheSumatratrenchandfrontal

subduction alongthe JavaTrench.Thisisaccompanied bythe ro-tation of the principal compressional stress direction, which is

oriented N–S west of the NER in the Central Indian Basin,

ex-plaining the formation of folds andE–W oriented reverse faults in this area (Bull and Scrutton, 1992), and NW–SE east of the NER in the Wharton Basin (Fig. 1b), where left lateral strike-slip faulting isprevalent along N–Strending re-activated fracture zones (Deplus et al., 1998). Therefore, the NER has been sug-gested toact asa naturalmechanicalbarrier separatingtwo dif-ferentdeformation patterns(DelescluseandChamot-Rooke,2007; Sager etal., 2013).However, therecentregionalstress field com-putation(GordonandHouseman,2015) indicatesthat thechange in the directionofmaximum compressive stress is moregradual fromN10◦Wat5◦NattheNER,increasingtoN20◦Watthe equa-torandN30◦Wfurthereastat95◦E(Fig.1b).

Ourstudyarealieswithin thenorthernWhartonBasin, which isboundedbytheSundaTrenchinthenorth,theNERinthewest and the fossil Wharton Spreading Centre in the south (Fig. 1b). The sediment thicknessinthe southern partofthe studyarea is about2.2 km(Singhetal.,2017).ClosertotheSundatrench, off-shore north Sumatra,the sediment thicknessranges between2.5 and 4 km(Moeremans etal., 2014). This sediment thickness re-flects the

>

200 meters per million years Nicobar fan deposition thatbegan

∼

9.5millionyearsago(Hüpersetal.,2017).

The Indian Oceanwithin its equatorialband displays a higher seismicity rate than any other ocean basin (Fig. 1b),

particu-larly afterthe2004Sumatra–Andamanearthquake (Wisemanand

Bürgmann,2012).SeismicityintheWhartonBasinandontheNER is dominatedby strike-slipsolutions (Fig. 1b),witha component ofthrustinginsomecases.ActivedeformationwithintheWharton Basin hadlongbeen detected butits true intensitywas revealed throughthe2012earthquakesequencesandmorerecentlybythe Mw 7.8strike-slipearthquakeonMarch2,2016(Layetal., 2016) furthersouth(Fig.1b),andwhichmakes itauniquesitetostudy intra-oceanicearthquakesandassociateddeformation.

3. Results

The high-resolution multibeam data were acquired using the

EM 122 MultibeamEcho Sounderproviding 5–10 mvertical and

25–50 m lateralresolutions and thesub-bottom 3.5 kHz profiles

using a Kongsberg SBP 120 system. The data were acquired at

N5◦Eazimuth, paralleltotheazimuthofthefracture zones,with

length varying from 100 to 400 km. Fig. S1 shows the location

of profiles. The earthquake locations were taken from National

Earthquake InformationCenter(NEIC) (http://earthquake.usgs.gov/ earthquakes/eqarchives/epic)andthefaultplanesolutionfromthe

Global Centroid Moment Tensor (GCMT) (http://www.globalcmt.

org/CMTsearch.html).

3.1. Generalmorphology

Bothun-interpretedandinterpretedmultibeambathymetry im-age coveringa region ofroughly 4.5◦

×

3.5◦ areshownin Fig.3.

Seafloor is relatively flat (4500 m), except near the NER and a

seamount discovered east of the NER. The main seamount rises

to 2600 m, and is elongated in ENE direction. Near the subduc-tion front,thetrenchischaracterizedby abathymetrylow,

rang-ing from 4500 m in the northwest to 4700 m in the northeast.

At about 100 km fromthe trench, one can clearly see a NW–SE

trending200 kmwideouterrisebulgeregion,associatedwiththe bending andflexureofthe downgoingoceaniclithosphere, punc-tuated by subsidence around the re-activatedN–S fracture zones andshearzones.Theflexuralbendisclearlyvisibleonthe3.5 kHz profile (Fig. 3c), whichalso showsextensivefaulting. The second most prominentfeatures are pervasive 290◦ striking shearzones

N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 177

Fig. 3. (a)Bathymetrymap(shadedreliefwithilluminationfromtheNE)oftheWhartonBasin,acquiredduringtheMIRAGE1cruise.Blackboxesshowlocationoffigures detailedinthefollowingsections.ThethickblackdashedlinesindicatereactivatedfracturezonesF6toF8outsideofourstudyarea.Whitedotsaremagneticanomaly picksandthewhitedashedlinesrepresentmagneticlineations.GreydotsanddashedgreylinesindicatethemagneticanomalyfromJacobetal. (2014).Smallblackcircles representlocationoftheIODPLeg362boreholes.Thecombinedred–greenarrowsindicatethedirectionsandtherelativemagnitudesoftheprincipaldeviatoriccompression andtensioncomponentsofstress(GordonandHouseman,2015).Pinkdashedlineindicatesthelocationof3.5 kHzprofileshowninfigure c.(b)Interpretedbathymetry showingN–Sre-activatedfracturezones(blackdashedlines),WNW–ESEshearzones(reddashedlines),NNE–SSWshearzones(yellowdashedlines),thrustsandnormal faults(black).BeachballsshowhistoricalseismicityrecordedintheWhartonBasinsince2010.Focalmechanismsrepresentallmagnitudes>5GlobalCMTsolutions(Ekström etal.,2012) fromJanuary2010toMay2017.(c)3.5kHzecho-sounderNorth–Southprofile 5(seealsoFig. S1forthelocation)showingtheflexuralbulge,intensefaulting aroundF6a,bendfaultsnearthetrench,andshearzones.Seafloorismarkedinlightblueandasedimentarylayerinyellow.

betweenthe re-activatedfracture zones.We havealsodiscovered newN20◦Estriking shear zones betweenthe fracture zonesand theNER.

The northeastern part of the study area gently dips toward thetrench andis marked by a series of sub-parallelgraben fea-tures,orientedroughly parallelto thetrenchdirection, relatedto bendfaulting.Arelativebathymetrichighisobservedintheregion northofthe seamount,whichisduetothrust faulting (discussed below).Towardthe southeastdomain,a large bathymetriclow is easilydistinguishable,forminganextensionalbasin.

Severalnorth–southorientedchannelsarepresentinthewhole study area; the most prominent channel is located just east of

the NER (Fig. 3a), flowing southward and can be observed over

170 kminourimage butcould betracedfurthernorthforabout 100 kmonpreviousbathymetry.The originofthischannel is dif-ficultto decipher asitliesbetweenthe NERandthe trench(see

alsoGeersenetal.,2015).Therearetwootherabandonedchannels furthereast,suggestingawestwardmigrationofthesechannels.

Inthefollowingsections,wewilldiscussthemaintectonic fea-turesobservedinourstudy:

3.2. Reactivatedfracturezones

ThenewbathymetrydatarevealeightN–Strendingre-activated fracturezones(Fig.3),F5b,F6a,F6b,F7a,F7b,andF8(Singhetal., 2011) from east to west. These features are typically orientated N8◦E(Fig. 4e)andshow a westwarddipcomponent. Sub-bottom 3.5 kHzprofilescrossingthefaultsshowverticaloffsetsofthe up-permostsedimentarylayersacrossthefracturezonesdocumenting recentactivity(Figs.3c,S2).

These faults are segmentedby enechelon compressional and extensional relays (Figs. 3b), common along strike-slip faults

Fig. 4. ShadedreliefbathymetryandtectonicinterpretationofthereactivatedfracturezonesF6initssouthernsegment(panelsaandb)andnorthernsegment(panelsc andd)(seeFig.3aforlocation).Theredbeachballrepresentsthefocalmechanismofthe2012mainshock.Blacksolidlinesshowseafloorfaultsandwhitearrowsindicate senseofmotion(sinistral).Whitedashedlineindicatesthelocationof3.5 kHzimageshowninFig. S2.Beachballsshowhistoricalseismicityinthebox.Yellowdotsindicate aftershocksbetween10Januaryand6November2012.(e)Rosediagramshowingthefrequencyoforientationofreactivatedfracturezones(12samples).

(Peacock andSanderson,1995). Theshape ofthe relays indicates

left-lateral movement, consistent with the focal mechanism of

earthquakesoccurringalongthesefeatures(Fig.3a).

Fracturezone F6aisthe mostprominentN–S strikingfracture zone,imaged hereovera 400 kmdistance (Fig.3a). F6a displays avariablesurfaceexpression.Inthesouth,itcorrespondstoaset ofpull-apart basins(Figs. 4b),characterized by vertical offsets of 10–80 mwithrespecttothesurroundingseafloor(Fig.5),and av-eragelengthsandwidthsofaround 5 kmand3 km,respectively. These pull-apart basins are formed as a result of transtension acrossreleasingstep-overs(Fig. 5). Tothenorth,F6aisexpressed byasetofenechelonnormalfaults(Fig.4d,Fig. S2).Weinterpret thesestructuresastensilecracksthatalsoindicatesinistralmotion alongtheN8

±

3◦E(Fig.4e)re-activatedfaults.Itisinterestingto notethatthesenormalfaultsbecomemuchmorepervasiveinthe outerriseregion,likelyduetotheeffectofplatebending(Fig. S2). TheepicenteroftheMw

=

8.62012earthquakeislocatednearthe F6afracturezone(Fig. 4c, d),suggestingthattheearthquake rup-turedfracturezone F6aasitsprimary N–StoNNE–SSW trending subfault. The epicenterofMw

=

7.2foreshock islocated slightly totheeast.FracturezoneF6brunsparalleltoF6aat

∼

12 km fur-therwestanddips eastward.Thissuggeststhat F6aandF6bmay beaconjugatepairoffaults,possiblyconnectingatdepth.

Anothermajorre-activatedfracturezonecanbeobservedwest ofF6:fracturezoneF7,whichshowsacomplexconfigurationwith two mainfaults (F7a, F7b).Deformation along thesefeatures

ap-pears diffuse with a dense network of small and very close en

echelonnormalfaults. IncontrasttoF6,pull-apart basinsare ab-sentalongthosesegments.TheepicenteroftheMw

=

8.2

earth-quake is located on the southern branch ofthe second segment

of this fracture zone (F7b; Fig. 3b), suggesting that the Mw

=

8.2earthquakemighthaverupturedF7b(Singhetal.,2017).Most of earthquakes in the region lie along the F6–F7 fracture zones (Fig.3b).Takentogether,theF6andF7re-activatedfracturezones formamajorstructuralboundary.

Thenewbathymetry dataalsoimage anewfracture zone, re-ferred to here as F5b, which had not been previously identified

duetothescarcityofmarinemagneticanomalydata(Jacobetal., 2014).Thisfaultisexpressedbyasetofenechelonnormalfaults in its northern part. In its southern part, deformation is much more intense, with a small extensional basin consistingof small grabens imagedalongthefault(Fig. 3b).Inthewestofourstudy area,re-activatedfractureF8isencountered,boundingtheNERon its eastside.Thestrikeoftheenechelonnormalfaultsalong the fracturezonesis330

±

5◦ (Singhetal.,2017).

Some ofthesere-activatedfracturezoneshavebeenidentified bothfurthernorthandsouthofourstudyarea.Usingbathymetry dataalong with3.5 kHz profiles,Graindorgeetal. (2008) showed that the F6 fracture zone reachesand intersects the deformation front.Cartonetal. (2014) haveimagedthere-activatedF5,F6and F7 fracture zoneson seismic reflectionprofiles, andshowedthat someofthemhaveanoffsetof300 matthebasement.Singhetal. (2017) havealsoshowntheseismicimagesofF6andF7withinthe sedimentsdowntothebasementatthesouthernextremityofour studyarea.Deplusetal. (1998) havealsoimagedthesere-activated fracture zones further south (5◦S) of our study area, suggesting thatthesefeaturesextendoverathousandkilometers.

3.3. Normalfaults

The northeasternpartofthesurveyedareaischaracterized by abundant small graben-type features (Fig. S3a) bounded by nor-mal faults (Fig. 6b). They are active and result from extensional bending stresses near the trench. Active normal faults are fre-quently observed affecting the oceanic plate topography in the vicinity of subduction zone trenches, such as in Central Amer-ica (Ranero et al., 2003). Such normal faults are induced by the flexure of theoceanicplateoutboard ofthe trench, andthey are

sometimes accompanied by normalfault earthquake focal

mech-anisms, such as those observed in Fig. 1b (green beach balls

near

∼

93◦E, 3◦N). The outer rise faulting may involve the re-activation ofthe oceanicspreadingfabric ofthe subductingplate if it is favorably oriented (Masson, 1991; Ranero et al., 2003; Grevemeyer etal., 2007). Here thestrike ofthe observednormal

N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 179

Fig. 5. (a)Un-interpretedand(b)interpretedbathymetricimageofapull-partbasinalongthesouthernsegmentofF6a.Blacklineindicatesthelocationofbathymetry(c) and3.5 kHzprofiles(d).Notethequalityandaccuracyof3.5kHzimage.Seafloorismarkedinlightblueandasedimentarylayerinyellow.

faults(N330◦E,Fig.6e)ismore orlessparallel tothe trenchand atanangletobothoceanicfracturezonesandabyssalhillnormal faults,suggestingthatthesenormalfaultsarenewlyformedfaults. In the southeast corner of the surveyed area, N–S oriented graben-typefeaturescanbeobserved(Fig.6c)southofthenewly discovered fracture zone F5b. Theyare not associated with plate bending,butlikelycorrespondtothesouthernterminationofF5b,

causing a wide extensional zone. These normal faults have the

samestrike,N330◦E(Fig.6e),asthatofthenormalfaultsnearthe subductionfront.Thedipofthesenormalfaultsis60–70◦ nearthe surface.

3.4.WNWshearzones

The majorityof the studyarea is cut by a set of right-lateral shearzones(Fig. 7),strikingatN290

±

5◦E(Fig. 7c). Theyappear asseries of graben-type features bounded by sets ofen echelon normalfaults strikingat N335

±

5◦E(Fig. 7d). Theseshear zones are brittle shear zones, since the deformation is concentrated in anarrow fault zone.Theyrepresentthe dominantstyle of defor-mationbetween there-activated fracturezones. The area located between the F5b and the F6a fracture zones is particularly af-fectedby severalshear zones parallel to each other spacedat 3 to6 km, distributed unevenly in the northernand inthe

south-ern parts. They change slightly their direction in the vicinity of F6a. Theseshear zonesare alsopervasive closeto theseamount, andcutthroughtheseamount(Fig.7b),buttheyareabsent north-westofthesystemofthrustfaults.Inthenorthwesternpartofthe studyarea,theseshearzonesagainbecomepervasive.

Towardthe western partof the studyarea (Fig. 3b), the NER is cut by a set offaults, which we interpret asWNW–ESE shear

zones, asthey are orientedWNW–ESE butalso becausethey are

consistentwiththeobserveddominanceofstrike-slipfocal mech-anisms for the northern NER earthquakes (Fig. 1b). Sager et al. (2013) have observed similar features furthernorth ofour study area.

Singh et al. (2017) discovered these shear zones using only limited data, but our results show that these shears are indeed veryprevalent.Theysuggestedthatthe2012greatWhartonBasin

earthquakes rupture proceeded in en echelon fashion with this

suite of N290◦E striking shear zones connecting the re-activated fracturezones(F6andF7),withanotherN–Strendingre-activated fracturezoneontheNER.Furthermore,inthesouthernpartofthe surveyed area just close to the seamount (between 91◦30E and 92◦10E,Fig.7b),one oftheWNWshearzonesseemstobe more developed,withaclearfreshtraceon thebathymetry,showinga largepull-apartbasinmanifestedbyenechelonnormalfaultsand boundedbynormalfaults(Fig. S4).Theshearzonealsooffsetsthe

Fig. 6. Shadedreliefbathymetrymap(a)andtectonicinterpretation(b)ofbending-relatednormalfaults.Beachballs showhistoricalseismicityinthebox.Yellowdots indicateaftershocksbetween10Januaryand6November2012.Shadedreliefbathymetrymap(c)andtectonicinterpretation(d)ofenechelonnormalfaultsofthesouthern terminationoftheF5breactivatedfault(seeFig.3aforlocations).Blacksolidlinesshowfaults,whitearrowsindicateF5b.Whitedashedlinesindicatethelocationsof3.5 kHz imagesshowninFigs.S3a(bendfaults)andextensionalfaults(Fig.3b).(e)Rosediagramshowingtheorientationofthesetwosystemsofnormalfaults(36samples). seamountbyapproximately2.5 km(Fig.7b).Furthermore,the

ma-jorityofaftershocksare locatedalongthisshearzone, suggesting that thisarea mighthave ruptured during the 2012great earth-quakesequence,possiblythefirstdirectseafloorevidence confirm-ingtheexistenceofaWNW–ESEcomponentofruptureduringthe Mw8.6event.

3.5. NNEshearzones

Westoffracturezone F7,particularlyatthewesterncornerof thebathymetry map, aseries ofcloselyspaced (5to 10 km) de-pressionscanbeobserved(Fig.8a).Thesedepressionscorrespond tograben-typefeatures, andformlongparallellineamentswitha NNE–SSWtrend.Weinterpretthesestructuresasleft-lateralshear zones(Fig. 8b)striking atN20

±

2◦E(Fig. 8c).Thesefeatures are similartotheshearzonesdescribedpreviouslyandhavethesame characteristicsasthetrendofenechelonfaultsboundingthesmall depressions (N335◦E) (Fig.8d). Furthermore,in thenorthwestern corner(

∼

92◦E),theyintersecttheWNW–ESEtrendingshearzones andthereactivatedfracturezones(at

∼

92◦30E)(Fig.3b).Theyare presentin thevicinity of the re-activatedfracture zones (F7 and F8), notin themiddle,and havethesame sense ofmotion, sug-gestingthattheyaregeneticallyrelated.

These shear zones are also brittle, since the deformation is concentrated in a narrow region, characterized by closely-spaced

faults, numerous graben basins and shear fractures. The small

grabens are bounded by enechelon faults; all these en echelon faultsshowapredominantlynormalsenseofmotionwitha verti-caloffsetontheseaflooronthescaleof5to10 m,withastrikeof 330

±

5◦(Fig.8d).Usingverylimitedbathymetrydata,Geersenet

al. (2015) noticedthesefeatures,andinterpretedthemasPRiedel shears.Asweshalldiscussbelow,theyarenotPRiedelshears, in-steadtheyareRRiedelshearsandareoneelementofthecomplex deformationintheWhartonBasin.

3.6. Thrustfaults

Following the work of Deplus et al. (1998), it has been

ac-cepted that the deformation in the Wharton Basinis

accommo-datedmainlybyN–Sstrikingre-activatedfracturezones.However, a minorityofthrustfaultmechanismsisrecordedinthesouthof theWhartonBasin(Deplus,2001),butnoevidenceofthrustfaults was directlyobservedontheseafloorintheavailablebathymetric data.

In thecentral partofourbathymetry map(92◦E,1◦30N), we observe upwarpedsediments indicatingdeformation along thrust faults (Fig. 9). Thesefaultsare presentmainlyinthearea located betweenfracture zonesF7andF8,closetotheNER.Theirlength rangesfrom10 kmupto45 kmandtheirdominantorientationis betweenN60◦EandN65◦E(Fig.9d).

Theyareclearlymarkedonthebathymetrybyhighscarpswith throw of 15–20 m (Fig. 9c), dipping 30◦ toward the northwest. Both on thewestern andeastern sides of thesethrusts,one can observeabandonedchannels,suggestingthatthesethrustsare re-cent.Themainactivechannellieswestofthesethrusts(Fig.3).

Sub-bottom 3.5 kHz profiles crossing these features (Fig. 8c) showhangingwallsup-thrustrelativetofootwalls.Moreover,they affect even the mostrecently deposited sediments and thus ap-peartobeactive.Onethrustearthquakeisobservedwestofthese thrusts,otherwisetheareaisdevoidofearthquakes(Fig.9b).

N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 181

Fig. 7. (a)Shadedreliefbathymetrymapand(b)tectonicinterpretationofWNW shearzones(seeFig.3aforlocation).Blacksymbolsshownormalfaultsbounding thesefeatures andwhite arrowsillustratesenseofmotion(dextral).Beachballs showhistoricalseismicityinthebox.Yellowdotsindicateaftershocksbetween10 Januaryand6November2012.Yellowdashedlineindicatesthepossiblelocation oftherupture.(c)and(d)Rosediagramsshowingthefrequencyoforientationof WNWshearzones(16samples)andrelatedenechelonnormalfaults(12samples), respectively.Whitedashedlineindicatesthelocationof3.5 kHzimageshownin Fig. S4.

Atthesouthwesternextremityofthissuiteofthrusts,thereis a30 kmlongand10 km wideseamount,orientedatN75◦E, sug-gestingthattheremightbealinkbetweenthepresenceofthrust faults and that of this seamount. West of F8, on the NER, there is some evidence of thrusting, with strike similar to the above thrusts.Sager et al. (2013) have noticed upwarping of sediment layers on seismic reflection profiles acquired farther north along theNER, indicating faults with compressionalmotion,striking at N100◦E,lessabundantthantheWNW–ESEshearzones.

These thrusts are orthogonal to the NW–SE maximum

com-pressivestressdirection(GordonandHouseman,2015).Theymay

indicatetheinitiationoffoldingandbucklingofthelithosphereas observedwestoftheNER.

4. Discussion

4.1. Linkbetweendifferenttypesoffaultingandregionalstressfield

Fig.10summarizestheorientationofdifferentfaults andtheir link withthelocal stressdirections. Basedon theorientations of allthesefaults,wecandefinetwodifferentstressdirections:

max-imum compressive andmaximum extensive stress axes oriented,

respectivelyN335◦EandN65◦E,orthogonaltoeachother,withan uncertainty of

±

5◦ (Fig. 10a). Strike-slip faulting occursin a

tri-axial stress field inwhich themaximum andminimumprincipal

stresses

σ

1and

σ

3layinthehorizontalplaneandtheintermediate principal stress

σ

2 is the vertical axis,thus the N335◦E oriented compressive stress mustbe theprincipal compressionalstress

σ

1 and the N65◦Eoriented extensive stress should be the principal extensionalstress

σ

3.

Fig. 10a alsosuggests that sinistral anddextral faults striking atN20◦E andN290◦E,respectively, are developedin response to these two main stress axes. These N290◦E and N20◦E trending

shear zones have the same orientation and shear sense

(right-handed and left-handed shear senses,respectively) asthese two conjugate fault systems (Fig. 10a). Therefore, the N290◦E and N20◦Eshear zonescorrespond to conjugatestructures developed inresponsetothelocalstressaxes.

It is well known that in the casewhere an active strike-slip zonelieswithinanareaofcontinuingsedimentationatlowlevels of strain, theoverall simple shear causesa set of smallfaults to form.Thedominantset,knownasRshear,formsat10–20◦tothe underlyingfaultwiththesameshearsenseasthemainfault.They oftenformanenechelonandoversteppingarray synthetictothe main fault. The R shears are then linked by a second set, the R shear,which formsatabout70–80◦ tothe mainfaulttrace (Katz etal.,2004),andthesetwoshearsformaconjugatesystem.

Inourcase,thestrikeofthereactivatedfracturezonesisN8◦E,

whereasthestrikeoftheNNE–SSWandWNW–ESEtrendingshear

zonesis N20◦EandN290◦E,respectively. The anglebetween the twosets ofshearzonesandstrike-slipre-activatedfracturezones is12◦and78◦,respectively.Consequently,weinterpretNNE–SSW

shear zones as R Riedel shears and WNW–ESE shear zones as

R Riedel shears. These two sets of shear zones appear to be

formedinresponsetodeformation betweenthere-activated frac-turezones,suggestingthatmostoftheregional-scaledeformation istakingplacealongstrike-slipfaults.

On the regional scale, the direction and magnitude of

max-imum compressive and extensional stresses vary from

predomi-nantlycompressiveatthenorthwestcornertoWNWcompressive

andextensional stressesatthesoutheastcornerofourstudyarea (Fig. 3a). The presence of the Riedel shears andthrusts between thefracture zonesF7 andF8mightbe duetothisregional varia-tionofstresses.

The trend ofre-activated fracture zones is oblique to the di-rectionofregionalcompressivestress;presumably,thismismatch between stress and fault direction could be a result of oppor-tunistic fracturing along major inherited lithospheric discontinu-ities i.e.fracture zones withlarge ageand crustalthickness con-trast.Thelarge-scaleregionalstressesmightberesponsibleforthe re-activationofthesefracturezones,whereas thestress field gen-eratedbetweenthefracturezoneswouldinturnencouragethe de-velopmentofRiedelshears betweenthefracture zones(Fig. 10b). Normally Riedel shears are shallow features (Kim et al., 2003) along deep-rootedstrike-slipfaults.The distancebetweenF7and

Fig. 8. (a)Shadedreliefbathymetrymapand(b)tectonicinterpretationofNNEshearzones(seeFig.3aforlocation).Blacksymbolsshownormalfaultsboundingshearzones andwhitearrowsillustratethesenseofmotion(sinistral).Reddashedlinesandwhitearrows(dextralsenseofmotion)markWNW–ESEshearzones.NNE–SSWshearzones (sinistralsenseofmotion)arealsomarked;notethatthemorphologyoftheseshearzonesissimilartothatoftheWNW–ESEstrikingshearzones.Yellowdotsindicate aftershocksbetween10Januaryand6November2012.(c)and(d)RosediagramsshowingthefrequencyoforientationofNNE–SSWshearzones(5samples)andrelateden echelonnormalfaults(16samples),respectively.

anenechelonpattern,formorethanhundredsofkilometers, sug-gestingthattheseRiedelshearsarealsodeeplyrootedona litho-sphericscale.

4.2. Shearzones:oceanicfabricreactivationornewlyformed structures?

Oceanic fracture zones of the fossil Wharton Spreading Cen-trearere-activatedasleft-lateralstrike-slipfaultsintheWharton Basin;itdoesnotseemtomatterthattheyare obliquetothe re-gionalstress pattern. Thisindicates that deformation hasatleast partlylocalizedon ancientzonesofweakness.Pre-existing struc-turesthusseem toplayan importantrole inthepresent-day de-formationoftheWhartonBasin.

By contrast, the N290◦E shears strike at an angle of 20–25◦

with respect to the dominantly E–W oceanic fabric based on

the magnetic anomaly pattern (Fig. 3a), and therefore they are likely newly formed. Since no pre-existing NNE–SSW orientation ispresentintheWhartonBasinoceaniccrust,theNNE–SSWshear zonesarealsonewlyformedtectonicfeatures.

4.3. Thrustfaultsandexistingstructures

The observed thrusts have a strike of N60–65◦E and the

seamountisorientedinasimilardirection.Themagneticanomaly dataindicate that thestrike ofthe magneticlineation inthis re-giondepartsfromthegeneralE–WdirectionandisalsoN60–65◦E, suggesting a local complexity of the Wharton Spreading Centre.

The seamount could have been formed during the crustal

accre-tion process or mayhave developedafterwards; we donot have

any constrains on the ageof theseseamounts. Itseems that the observedthrustshavedevelopedalongsomeofthesepre-existing volcanicfabrics.

4.4. Effectofbending

The re-activatedfracture zonesandassociatedshearzones be-comemuch morepervasiveintheouter riseregionofthe

Suma-tran trench; this indicates a complex interaction between the

bending stress andthe principalcompressive stress. The bending

stress induces a 200-km wide flexuralbulge andabend towards

the subductionfront, withabundant extensional deformationand thus an enhanced role of normalfaults compared to farther sea-ward. For example, the re-activated fracture zone F6a is formed by closely spaced en echelon normal faults in the region of the bend (Fig. S2).The strike ofnormal faults remains the same,i.e. 330–335◦azimuthsaselsewhereinoursurveyarea.

The bending stresses are generally normal to the subduction direction, whichisobliquetothetrenchinthisregion.The obliq-uityleadstoaslippartitioningbetweentrench-orthogonalmotion

along the megathrust and trench-parallel motion accommodated

by strike-slipfaults intheforearc region(Fitch,1972).Toaccount forpossiblerupturingbythe2012Mw

=

8.6earthquakeofafault segmentsubparalleltothetrench,Ishiietal. (2013) suggestedthat trench parallel strike-slip faulting due to slip partitioning could alsooccur intheincomingplateseaward ofthetrench.However,

N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 183

Fig. 9. (a)Shadedreliefbathymetrymapand(b)tectonicinterpretationofthrust faults(seeFig.3aforlocation).Blacksymbolsshowreversefaults.Dashedwhite lineindicatesthelocationofsub-bottom3.5 kHzprofileshowninfigure c.Beach ballsshowhistoricalseismicityinthebox.Yellowdotsindicateaftershocksbetween 10Januaryand6November2012.Theblacklinesindicatethrustfaults.(d)Rose diagramshowingthefrequencyoforientationofthrustfaults(3samples).

the WNW-striking shear zones are very pervasive and extend at

least500 km seaward of the trench, too far forthem to be the seaward expression of slippartitioning; these shears are instead causedbythedeformationmechanismexplainedinFig.10.Onthe other hand, as the strikes of bending related normal faults and

those forming the WNW shear zones are the same, it is rather

difficultto separate the effect ofbending-related stress fromthe regionalstress.

4.5.Linkbetweentheactivefaultingandthe2012earthquakes

Manyof the 2012 published earthquake models indicate that

the Mw 8.6 main-shock involved rupture on multiple NNE–SSW

andWNW–ESEtrendingfaults(Ishiietal.,2013; Mengetal.,2012; Yueetal., 2012; Wei etal.,2013), andwhilethesegment show-ingthe largestfractionofseismicmoment releasewassuggested

to be either NNE–SSW trending or WNW–ESE trending, our

ob-Fig. 10. (a)Rose diagramshowing thestrikesofallnormal(orange)and thrust (blue)faultsdefiningthedirectionsoftheintra-platecompressivestressσ1(inward

pointingredarrows)andtheextensivestressσ3(outward-pointinggreenarrows).

TheRiedelshears(RandR)(red:WNW–ESEshearzones,green:NNE–SSWshear zones)indicatetheconjugatefaultsystems,withblackarrowsindicatingthesense ofdisplacement.(b)Schematicdiagramsummarizingregionalandlocal deforma-tionregimeshowingallthefaultsandstresses.Mostofthecompressivestresses aretakenalongthere-activatedfracturezonesandtherestalongtheshearzones.

servationsofa prominentF6atendto agreewiththeformer.The Mw

=

8.2 aftershock occurred two hours later, rupturing a sec-ond NNE–SSW trending fault. Besides the fact that spatial corre-spondenceofearthquakesonoceanicplateswithmappedfaultsis oftendifficultto ascertaindueto locationinaccuracies, itis clear that the Mw

=

8.6epicenter aligns along theF6a fracture zone.

For the Mw

=

8.2, the epicenter aligns along the F7b fracture

zone. We suggest that the Mw 8.6 earthquake ruptured several

distinct fault segments, starting fromthe F6a and ending atthe

F7b.TheNNE–SSWshearzoneslocatedbetweenthenorthernand

thesouthernsegmentsoftheF7bfracturezonelikelylinkedthose segmentsandtransferredrupturealongtheminenechelonform, andthus,mayhavehostedsomeofthemomentreleaseduringthe

Mw

=

8.6earthquake.Otherwise,some slipmighthaveoccurred

along oneoftheseshears.This explainstheNEtrendcomponent

shown by some of the earthquake models (Meng et al., 2012;

Weietal.,2013).Furthermore,theN290◦Estrikingshearzone lo-catedclosetotheseamountisthebest candidateforhosting the WNW–ESE rupture(Fig. 7b);the dense patternof N290◦E shear zonesinthisareastronglysuggestshighstressdropalongseveral structuresinenechelonform.Anumberofaftershocksalsoalign thisshear zone (Fig. 7b). The N290◦E shears are presentalso on theNER,andtheylikelycorrespondtotheWNW–ESEbathymetric lineations mapped bySager etal. (2013). We suggest that WNW

shearzonesconnect therupturedre-activatedfracturezones(F6a andF7b) withanotherN–Strendingre-activatedfracturezoneon thewesternsideoftheNER.

Our findings differ in detail from existing proposed rupture models, due to the fact that the inversion of seismological and geodetic data record can only use long wavelength signals hun-dredsandthousandsofkilometersawayfromtherupturelocation. Consequently, thesedata cannot resolve the slip along each seg-mentseparately; instead,they provideanaverage solution,which toalargeextentpredicttheseafloorobservations.

The April 2012 earthquakes have their centroids at 45.6 and 54.7 kmdepth,respectively(GlobalCMTCatalog).Furthermore, fi-nitefault modelsoftheseearthquakes suggestalsorupturedown tothebaseofthe lithosphereata depthof50to60 km(Wei et al.,2013; Yueetal.,2012).Almosttheentiredepthoftheoceanic lithospherehasruptured.QinandSingh (2015) haveimagedfaults downtoadepthof45 kminthenortheastoftheWhartonBasin, supportingtheidea oflithosphericscalefaulting duringthe 2012 earthquake(Weietal.,2013).

4.6. TransitionfromN–Sstrike-sliptothrustfaultingacrosstheNER

Our bathymetry data show the presence of WNW–ESE shear

zoneswest of the F8, similar to the Wharton Basin. Sager et al. (2013) alsoobservesimilarshear zonesfurtherwestonthe NER. Theexistence ofstrike-slipearthquakeson theNERsuggeststhat theseshearzonesareactive.FurtherwestoftheNERinthe Cen-tralIndianBasin,westofourstudyarea,mostoftheearthquakes havestrike-slip focal mechanisms(Fig. 1a). Althoughit hasbeen acceptedthattheN–Sthrustingisadominantprocesswestofthe NER(e.g.,BullandScrutton,1992),thepresenceoftheWNW–ESE shearzonesontheNERandthelargenumberofstrike-slip earth-quakeswestoftheNERsuggestthat thedeformationwestofthe

NER mightbe similar to the Wharton Basin, more complexthan

previouslyrealized,andmoredataarerequiredtoanswerthereal relationshipbetweenthrustingandstrike-slipfaulting.

4.7. Timesequenceofdifferentfaults

TheonsetofdeformationintheWhartonBasinisnotwell

con-strained. Using seismic reflection data and assuming an age of

40 Myr for the Nicobar fan sediments, Singh et al. (2017) sug-gested that the reactivation along the fracture zones started at 17.5 Myr ago.However, theresultsfromIODPdrillingsiteU1480 andU1481offshore Sumatra indicatethat Nicobarfan deposition began9.5 Myr ago (Hüperset al.,2017) and shut downat 2 Ma (McNeiletal.,2017),requiringthedeformationbemuchyounger. However,astheIODPboreholeslieinthevicinityoftheNERand onthebendingrelatedflexuralbulge,theseresultsmaynotbe ap-pliedtothewholeWhartonBasin,particularlytothefractureF6a, whichismostactive.

The sequential development of shear surfaces within a natu-ral Riedel system is difficult to ascertain. Thus, it is difficult to determine the age sequence of different fault systems, but gen-erallyspeaking,themostwidely acceptedmodelforRiedelshear zone development is synthetic driven in the sense that R Riedel shears are normally the first subsidiary fractures to occur and generally build the most prominent set (Bartlett et al., 1981; Katzetal.,2004).Butinourcase,the WNWshearzones(R)are moreprevalent.

Itiswellacceptedthatdeformationgoesthroughexistingareas ofweakness.Since thefracture zonespresentzonesofweakness, stressvectorsstartfirstby reactivatingthemasleft-lateral strike-slipfracturezones.Asthesefracturezonesareobliquetoprincipal stressvector(orientedN335◦E),deformationcreatednewRRiedel

shear zones to accommodate the rest of strain between the

re-activatedfracture zones. Inordertoaccommodate therotationof these R shear zones, a second network of R Riedel shears ori-entedWNW–ESEdeveloped.Iftherelativemotionalong neighbor-ingfracturezonesisdifferent,itcouldaccentuatethedevelopment of R Riedel shears, and lead to lithospheric scale shear zones.

Some of these shear zones extend up to the uplifted area due

to thrusting north of the seamount, suggesting that either shear

zones are youngest or are the most dominant type of

deforma-tionbetweenthefracturezone.However,thepresenceofinactive channels in this uplifted area indicates that they have migrated westwards, suggesting that thesethrust faults are still active. Fi-nally, bending-relatedfaults are the mostrecently formed, since they aretheresultofoceanicplatedeformation duetothe bend-ingjustpriortosubduction.

5. Conclusions

High-resolutionmultibeam bathymetry,alongwithsub-bottom

profiler data from the source region of the 2012 Wharton Basin

earthquake,revealactive processesshapingtheseafloor.Basedon thesenewdata,thefollowingconclusionscanbedrawn:

1. Eight re-activated left-lateral N8◦E striking fractures (F5b,

F6a, F6b, F7a, F7b and F8) are present in the 2012 earthquake

rupture zone area. F5b is a newly discovered fracture zone, ter-minatedbyanextensionalbasinatitssouthernend.Fracturezone F6a appearstobethemostactiveone andhasbeenimagedover adistanceof450 kminthenewdata;itmostlikelycontinues to-ward the subduction front inthe north andextends all the way tothefossilWhartonSpreadingCentreinthesouth(Deplusetal., 1998).Thismajorstructurelikelyrupturedduringthe2012Mw

=

8.6earthquake.

2.RightlateralN290◦ strikingshearzones, presentinbetween the above re-activatedfracture zones(including onthe NER), are

the most pervasive seafloor structures. They are 2–3 km wide,

formedbysetsofenechelonnormalfaults,andstrikeobliquely(at a20◦angle)totheE–Wridgefabricbasedonmagneticanomalies, suggestingthat theyarenewly formedstructures. IntheSWpart

of the survey area, one such prominent shear zone crosses and

offsets a large seamount by 2.5 km. This shear zone is the best

candidateforanESE–WNW componentoftheruptureduring the

2012Mw

=

8.6earthquake.

3. N20◦E striking left lateral shear zones are imaged west of fracture zoneF7; theyappear similartothe N290◦ strikingshear zonesbutareorientedorthogonaltothem.Theirstrikeisin

agree-ment with the strike of one of the nodal planes of the GCMT

solution of the Mw

=

8.6 event, but it is not clear ifthey rup-turedduringtheearthquake.

4. En echelon N335◦ striking normal faults are ubiquitous in thenorthernWhartonBasinandformkeyelementsfortheabove threetypesofstrike-slipfeatures.Inaddition,theyalsoboundan extensional basinsouthofF5bandaccommodatebendingrelated faultinginthenorth.

5.ThrustfaultsareimagedbetweentheNERandfracturezone F7strikingN65◦E,andtheir orientationisorthogonaltothe max-imum directionof compressionin theregion. Thesethrust faults are parallel to thelocal magneticanomalies, andare boundedin thesouthbyanelongatedseamountdiscoveredduringthe experi-ment,suggestingthattheyareassociatedwithpre-existingcrustal fabric.

6. The strikes of the normal and thrust faults, orthogonal to

each other, define the directionof maximum compressional and

extensionalstressesintheregion,respectively.Ontheotherhand, theN20◦EandN290◦strikingshearzones,alsoorthogonaltoeach other,defineRandR Riedelshears,respectively,whichcombined with the N8◦E re-activated fracture zonesexplain thewhole

de-N. Hananto et al. / Earth and Planetary Science Letters 502 (2018) 174–186 185

formation pattern in the region. The stresses localized along re-activated fracture zones, particularly F6a, seem to accommodate

most of the N–S component of deformation, with some

discon-tinuousanddistributeddeformationaccommodated bytheN20◦E strikingshear zones. Theother component ofdeformation seems tobe distributedalong thepervasive N290◦ strikingshear zones, mayberecentlylocalizedalongtheWNWshearzonethathas pro-ducedanoffsetof2.5 km.

Acknowledgements

ThedatawereacquiredonboardtheIPEVR/VMarionDufresne, undera collaboration between the IPEV, Institut de Physique du Globe de Paris, Earth Observatory of Singapore, and Indonesian InstituteofScience. Wewouldto thankClaudio Satrianofor pro-viding re-locatedaftershocksshown inFigs. 4–9.Thisresearch is supportedbytheNationalResearchFoundationSingaporeandthe

Singapore Ministry of Education under the Research Centres of

Excellenceinitiative.ThisworkcomprisesEarthObservatoryof Sin-gaporecontributionno. 215andInstitutdePhysiquedu Globede Pariscontributionnumber3972.

Appendix A. Supplementarymaterial

Supplementarymaterialrelatedtothisarticlecanbefound on-lineathttps://doi.org/10.1016/j.epsl.2018.09.007.

References

Amante,C.,Eakins,B.W.,2009.ETOPO1GlobalReliefModelConvertedtoPanMap LayerFormat.NOAA-NationalGeophysicalDataCenter.

Bartlett,W.L.,Friedman, M.,Logan, J.M.,1981.Experimental foldingandfaulting ofrocksunderconfiningpressure:partIX.Wrenchfaultsinlimestonelayers. Tectonophysics 79,255–277.

Bull,J.M.,Scrutton,R.A.,1990.FaultreactivationinthecentralIndianOceanandthe rheologyofoceaniclithosphere.Nature 344,855–858.

Bull,J.M.,Scrutton,R.A.,1992.Seismicreflectionimagesofintraplatedeformation, centralIndianOcean, and theirtectonicsignificance.J. Geol.Soc.Lond. 149, 955–966.https://doi.org/10.1144/gsjgs.149.6.0955.

Carton,H.,Singh,S.C.,Hananto,N.D.,Martin,J.,Djajadihardja,Y.S.,Udrekh,Franke, D.,Gaedicke,C., 2014.DeepseismicreflectionimagesoftheWharton Basin oceaniccrustanduppermostmantleoffshorenorthernSumatra:relationwith activeandpastdeformation.J. Geophys.Res. 119,32–51.

Delescluse,M.,Chamot-Rooke,N.,2007.Instantaneousdeformationandkinematics ofthe India–Australiaplate. Geophys.J.Int. 168,818–842.https://doi.org/10. 1111/j.1365-246X.2006.03181.x.

Delescluse,M., Chamot-Rooke,N., Cattin,R.,Fleitout,L., Trubienko,O.,Vigny,C., 2012.April2012intra-oceanicseismicityoffSumatraboostedbythe Banda-Acehmegathrust.Nature 490,240–244.

Deplus,C.,2001.IndianOceanactivelydeforms.Science 292,1850–1851.https:// doi.org/10.1126/science.1061082.

Deplus,C.,Diament,M.,Hebert,H.,Bertrand,G.,Dominguez,S.,Dubois,J.,Malod,J., Patriat,P.,Pontoise,B.,Sibilla,J.-J.,1998.Directevidenceofactivedeformation intheeasternIndianoceanicplate.Geology 26,131–134.

Duputel,Z., Kanamori, H.,Tsai, V.C.,Rivera, L., Meng, L.S., Ampuero, J.P., Stock, J., 2012. The 2012 Sumatra great earthquake sequence. Earth Planet. Sci. Lett. 351–352,247–257.

Ekström,G.,Nettles,M.,Dziewo ´nski,A.M.,2012.TheglobalCMTproject2004–2010: centroid moment tensors for 13,017 earthquakes. Phys. Earth Planet. In-ter. 200–201,1–9.https://doi.org/10.1016/j.pepi.2012.04.002.

Fitch,T.J., 1972.Plateconvergence, transcurrentfaults,and internaldeformation adjacenttoSoutheastAsia andthewestern Pacific.J. Geophys.Res. 77(23), 4432–4460.

Geersen,J.,Bull,J.M.,McNeill,L.C.,Henstock,T.J.,Gaedicke,G.,Chamot-Rooke,N., Delescluse,M.,2015.Pervasivedeformationofanoceanicplateand relation-shiptolarge>Mw 8intraplateearthquakes:thenorthernWhartonBasin,India Ocean.Geology 43,359–362.

Gordon,R.G.,2000.Diffuseoceanicplateboundaries:strainrates,verticallyaveraged rheology,andcomparisonswithnarrowplateboundariesandstableplate inte-riors.In:Richards,M.A.,Gordon,R.G.,vanderHilst,R.D.(Eds.),TheHistoryand DynamicsofGlobalPlateMotions.In:GeophysicalMonograph,vol. 121.AGU, Washington,DC,pp. 143–159.

Gordon,R.G.,Houseman,G.A.,2015.DeformationofIndianOceanlithosphere: evi-denceforahighlynonlinearrheologicallaw.J.Geophys.Res. 121,4434–4449.

Graindorge, D.,Klingelhoefer,F.,Sibuet,J.-C., McNeill, L.,Henstock, T.J.,Dean,S., Gutscher,M.-A.,Dessa,J.X.,Permana,H.,Singh,S.C.,Leau,H.,White,N.,Carton, H.,Malod,J.A.,Rangin,C.,Aryawan,K.G.,Chauhan,A.,Galih,D.R.,Greenwood, C.J.,Laesanpura,A.,Prihantono,J.,Royle,G.,Shankar,U.,2008.Impactoflower platestructureon upperplatedeformationat theNWSumatranconvergent marginfromseafloormorphology.EarthPlanet.Sci.Lett. 275,201–210.

Grevemeyer,I.,Ranero,C.R.,Flueh,E.R.,Klaeschen,D.,Bialas,J.,2007.Passiveand activeseismologicalstudyofbending-relatedfaultingandmantle serpentiniza-tionattheMiddleAmericaTrench.EarthPlanet.Sci.Lett. 258,528–542.https:// doi.org/10.1016/j.epsl.2007.04.013.

Heine,C.,Müller,R.D.,2005.LateJurassicriftingalongtheAustralianNorthWest Shelf:margingeometryandspreadingridge configuration.Aust.J.EarthSci.: Int.Geosci.J.Geol.Soc.Aust. 52(1),27–39.

Hill,E.M.,Yue,H.,Barbot,S.,Lay,T.,Tapponnier,P.,Hermawan,I.,Hubbard,J., Baner-jee, P.,Feng, L., Natawidjaja, D., Sieh,K., 2015. The2012 Mw 8.6 Wharton Basinsequence:acascadeofearthquakesgeneratedbynearorthogonal,young, oceanicmantlefaults.J.Geophys.Res. 120,3723–3747.

Hüpers,A.,Torres,M.E.,Owari,S.,McNeill,L.C.,Dugan,B.,Henstock,T.J.,Milliken, K.L.,Petronotis,K.E.,Backman,J.,Bourlange,S.,ChemaleJr.,F.,Chen,W., Col-son,T.A., Frederik, M.C.G., Guèrin, G., Hamahashi,M., House, B.M., Jeppson, T.N.,Kachovich,S.,Kenigsberg,A.R.,Kuranaga,M.,Kutterolf,S.,Mitchison,F.L., Mukoyoshi,H.,Nair,N.,Pickering,K.T.,Pouderoux,H.F.A.,Shan,Y.,Song,I., Van-nucchi,P.,Vrolijk,P.J.,Yang,T.,Zhao,X.,2017.Releaseofmineral-boundwater priortosubductiontiedtoshallowseismogenicslipoffSumatra.Science 356 (6340),841–844.https://doi.org/10.1126/science.aal3429.

Ishii,M.,Kiser,E.,Geist,E.L.,2013.Mw8.6Sumatranearthquakeof11April2012: rareseawardexpressionofobliquesubduction.Geology 41,319–322. Jacob,J.,Dyment,J.,Yatheesh,V.,2014.Revisitingthestructure,age,andevolution

oftheWhartonBasintobetterunderstandsubductionunderIndonesia.J. Geo-phys.Res. 119,169–190.

Katz,Y.,Weinberger,R.,Aydin,A.,2004.GeometryandkinematicevolutionofRiedel shearstructures,CapitolReefNationalPark,Utah.J.Struct.Geol. 26,491e501.

https://doi.org/10.1016/j.jsg.2003.08.003.

Kim, Young-Seog,et al., 2003.Mesoscale strike-slipfaults and damage zonesat Marsalforn,GozoIsland,Malta.J.Struct.Geol. 25,793–812.

Lay,T., Ye, L., Ammon, C.J., Dunham, A., Koper, K.D.,2016. The2 March 2016 Wharton BasinMw7.8 earthquake: highstressdropnorth–southstrike-slip rupture in the diffuse oceanic deformation zone between the Indian and AustralianPlates. Geophys.Res. Lett. 43, 7937–7945.https://doi.org/10.1002/ 2016GL069931.

Liu,C.-S.,Curray,J.R.,McDonald,J.M.,1983.Newconstraintsonthetectonic evolu-tionoftheeasternIndianOcean.EarthPlanet.Sci.Lett. 65(2),331–342. Masson,D.G., 1991. Faultpatternsat outertrench walls.Mar. Geophys.Res. 13,

209–225.

McGuire,J.J.,Beroza,G.C.,2012.ArogueearthquakeoffshoreSumatra.Science 336, 1118–1119.

McNeil,L.C.,etal.,2017.UnderstandingHimalayanerosionandthesignificanceof theNicobarFan.EarthPlanet.Sci.Lett. 475,134–142.

Meng,L.,Ampuero,J.-P.,Stock,J.,Duputel,Z.,Luo,Y.,Tsai,V.C.,2012.Earthquake inamaze:compressionalrupturebranchingduringthe2012Mw8.6Sumatra earthquake.Science 337,724–726.

Moeremans,R.,Singh,S.C.,McArdle,J.,Johansen,K.,2014.Seismicimagesof struc-tural variations along the deformation frontof the Andaman–Sumatra sub-ductionzone:implicationsforrupturepropagationandtsunamigenesis.Earth Planet.Sci.Lett. 386,75–85.

Molnar, P., Tapponnier, P., 1977. The collision between India and Eurasia. Sci. Am. 236,30–41.

Peacock, D.C.P.,Sanderson, D.J.,1995.Strike-slip relayramps.J. Struct. Geol. 17, 1351–1361.

Petroy, D.E., Wiens, D.A., 1989. Historical seismicity and implications for dif-fuse plate convergence in the northeast Indian Ocean. J. Geophys. Res. 94, 12301–12319.

Prawirodirdjo, L., Bock, Y., Genrich, J.F., Puntodewo, S.S.O., Rais, J., Subarya, C., Sutisna,S.,2000.OnecenturyoftectonicdeformationalongtheSumatranfault fromtriangulationandGPSsurveys.J.Geophys.Res. 105,28,343–28,361. Qin,Y.,Singh,S.C.,2015.Seismicevidenceofatwo-layerlithosphericdeformation

intheIndianOcean.Nat.Commun. 6,8298.

Ranero, C.R.,PhippsMorgan, J., McIntosh, K.,Reichert,C.,2003. Bending-related faultingandmantleserpentinizationattheMiddleAmericaTrench.Nature 425, 367–373.

Royer,J.-Y.,Gordon,R.G.,1997. ThemotionandboundarybetweentheCapricorn andAustralianplates.Science 277,1268–1274.https://doi.org/10.1126/science. 277.5330.1268.

Sager,W.W.,Bull,J.M.,Krishna,K.S.,2013.ActivefaultingontheNinetyeastRidge anditsrelationtodeformationoftheIndo-Australianplate.J.Geophys.Res. 118, 4648–4668.https://doi.org/10.1002/jgrb.50319.

Sager,W.W.,Fullerton,L.G.,Buffler,R.T.,Handschumacher,D.W.,1992.ArgoAbyssal Plain magnetic lineations revisited: implications for the onset of seafloor spreadingandtectonicevolutionoftheeasternIndianOcean.In:Proceedings oftheOceanDrillingProgram,ScientificResults,vol. 123.CollegeStation,Texas, ODP.

Singh,S.C.,Carton,H.,Chauhan,A.S.,Androvandi,S.,Davaille,A.,Dyment,J., Can-nat,M.,Hananto,N.D.,2011.ExtremelythincrustintheIndianOceanpossibly resultingfrom plume–ridge interaction. Geophys. J.Int. 184, 29–42.https:// doi.org/10.1111/j.1365-246X.2010.04823.x.

Singh,S.C.,Hananto,N.,Qin,Y.,Leclerc,F.,Avianto,P.,Tapponnier,P.E.,Carton,H., Wei,S.,Nugroho,A.B.,Gemilang,W.A.,Sieh,K.,Barbot,S.,2017.Thediscovery ofaconjugatesystemoffaultsintheWhartonBasinintra-platedeformation zone.Sci.Adv. 3,e1601689.https://doi.org/10.1126/sciadv.1601689.

Wei,S., Helmberger, D., Avouac, J.-P., 2013. Modeling the 2012 Wharton Basin earthquake off Sumatra; completelithospheric failure.J. Geophys. Res. 118, 3592–3609.

Weissel,J.K.,Anderson,M.S.,Geller,C.A.,1980.DeformationoftheIndo-Australian plate.Nature 287,284–291.

Wiens,D.A.,Stein,S.,1984.Intraplateseismicityandstressesinyoungoceanic litho-sphere.J.Geophys.Res. 89,11442–11464.

Wiseman,K.,Bürgmann,R.,2012.StresstriggeringofthegreatIndianOcean strike-slipearthquakeinadiffuseplateboundaryzone.Geophys.Res.Lett. 39,L22304.

https://doi.org/10.1029/2012GL053954.

Yue,H.,Lay,T.,Koper,K.D.,2012.Enechelonandorthogonalfaultrupturesofthe 11April2012greatintraplateearthquakes.Nature 490,245–249.